[R] Savitzky-Golay filtering with missing data

[Ryan Hope]

I have a timeseries with some missing data points that I need smoothed
using a Savitzky-Golay. Right now I am using the sgolayfilt function
in the signal package. It fails if I have NA values in the timeseries.
If I replace NA with NULL sgolayfilt does not fail but it drops the
NULL values which is not really a valid solution. Is there anything I
can do about this?

-- 
Ryan Hope, M.S.
CogWorks Lab
Cognitive Science Department
Rensselaer Polytechnic Institute